It has become a common theme in DNA research papers dealing with population genetics, particularly those that are either published in the "west" or else rely heavily on references to previous publications by "western" research teams, to dichotomize human phylogeny neatly into two main types: African and non-African. As a result, a good amount of the readership of these papers have also become accustomed to treating human phylogeny accordingly. No doubt that the reactionary segment of that readership have applied such a phylogenetic arrangement synonymously with "races" of humanity.
It's one thing to assign human phylogeny into two main types, but it's another, in terms of how these assignments take form. One would be hard-pressed not to come across a single example, whereby lineage that is given an "L" designation in mictochondrial phylogeny, is automatically treated as "African", while the two main offshoots of the L3 clade are taken for granted as "non-African". Such arrangements tacitly or by design, insinuate non-overlapping phylogeny between the so-named two main groups. The discussion section will deal with this kind of arrangement of human phylogeny further, applying specific examples from published material.
Only as recent as a month ago, potential observers on this site would have come across a quintessential example of the aforementioned arrangement of human phylogeny, suggestive of non-overlapping phylogeny between the so-named two main types (African and non-African), within the Pagani et al.'s (2012) research paper on Ethiopian "genetic diversity"  that was discussed. Therein, the authors automatically took anything that was considered to be an "L" designated mtDNA clade as "African", and took any mtDNA what was not designated as such, namely for example—the M and N offshoots of hg L3, gratuitously as "non-African".
Pagani et al.'s practice steers clear away from the real complexities of human phylogeny, by hinging their assignment effortlessly on what is considered an "L" clade and what is not, and thereof, facilitating assignment into two neat, seemingly non-overlapping, ancestral lines. The basis for the assignment of remaining segments of the genome, which were not mitochondrial, into an identical two-party grouping of ancestral lines (African and non-African respectively), was mostly left to the imagination of the reader. The few exceptions therein, where the DNA locus was implicated by the name of a gene, as was the case of the SLC24A5 gene, "frequency" was alluded to as the reasoning behind the assignment into one of the two camps of ancestry.
The "derived" variant of the SLC24A5 gene was taken for granted as "non-African" on the mere account of its high "frequencies" in European samples, but it was demonstrated  that the distribution pattern of the variant, along with the relevant attributes of other pigmentation genes, elicited caution against that assumption. Among other sub-Saharan groups, the "derived" variant of the SLC24A5 appears in the gene pool of the San, as have "derived" variants of other known "skin pigmentation" genes.
The easing up of skin eumelanin in San hunter-gatherers has generally been attributed to local evolution in lower UV radiation environments they frequent, as opposed to the result of gene flow. In the Ethiopian samples, on the other hand, the presence of the "derived" variant of the SLC24A5 gene was peculiar in that it was not found in tandem with other "skin-pigmentation" affiliated genes whose distribution generally paralleled that of the "derived" SLC24A5 variant, particularly in Europeans. Hence, "frequency" in itself is not a sufficient enough indicator for ascribing a single-source origin in the form of a "non-African" origin.
While many research papers from the "west" tend to take it for granted that such clades as M and N are "non-African", there is truly no such clarity that points to a non-African origin for either sub-clades of L3. Observations have appeared every now and then in research papers that put to question such unidimensional viewpoints of human phylogeny. As an example, one paper notes:
It is plausible that the ‘M and F’ lineage could have originated in East Africa and migrated to India through the southern route, but the alternative of a back migration to Africa cannot be ruled out at this point. The HpaI/HincII loss at np-12406 (characteristic of ‘F’) could have arisen from a mutation in any of the four nucleotides, which can lead to multiple origins of ‘F’.
Thus the African haplotypes AF21 and AF10 (Figure 2) could have arisen from an independent mutation (‘F’) in a ‘+DdeI 10394, –AluI 10397’ African lineage. These haplotypes have a ‘T’ at nucleotide 16223, whereas the Indian and East Asian ‘N and F’ have a ‘C’ at this position.
If the Asian haplotype originated from the one in Africa by the loss of the variable DdeI 10394 site, then the 16223T → C transition could have taken place either in Africa or India, as seen in Figure 2, or the Asian ‘N and F’ lineage could have arisen on an African ‘N’ background.
‘F’ is found at a higher frequency in East Asia, where it is well differentiated; its origin in this region would involve yet another independent event. - Barnabas et al. 2005 
Ideologues, who have a vested interest in making emotional points rather than pursue fact, often seize on the date of such publications, as the Barnabas et al. report could potentially exemplify, to avoid having to deal with implications of revelations that seem to question the veracity of "popular" viewpoints, when there is little else to take issue with, from a substantive standpoint. Such revelations would otherwise tacitly call upon a new direction to how human phylogeny is approached, which "traditionalists" may find harder to adapt to than more open-minded observers.
Though the cited Barnabas et al. work is about 8 years old, the pressing matters in it—with regards to the challenges that come with the pursuit of firmly pinning down the mentioned clades to a definitive singular origin—are still very much alive, notwithstanding some papers or the others purporting to have a firm grounding on the origins of these so-called "non-African" clades.
The Barnabas et al. notes above open up the possibility of an African origin for these clades, just as a few other studies have. Still, many others take a less objective approach and treat these clades as though the authors have firm answers on their origins; this is cause for readers and observers to pause and take caution: origins of many of these "non-African" clades is still far from certain, notwithstanding what some research papers may be implicitly or explicitly suggesting otherwise.
Let's take for instance, the above-cited observations that urged Barnabas et al. to consider an African origin for M and N clades, or say, the F sub-clade of N. It would appear that in applying the "conventional" nucleotide attributes (reference material, like the Cambridge Reference Sequence data) used for clade identification against their observations, the HpaI/HincII loss at np-12406 in the African haplotypes of AF21 and AF10 compelled Barnabas et al. to consider an African origin for the F clade, if not the possibility of a multiple-origin scenario.
The plausibility of "multiple-origin" scenario is another thing that seems to elicit many geneticists to put up their guards, even when it appears that could very well have been the case. For instance, consider the treatment of the np 12308G insertion as UEP, which reportedly characterizes the U clade and its sub-clade K. The appearance of the same substitution within the hg T family (Pala et al. 2012) , however, suggests that 12308A-G SNP may have emerged more than once. Also consider this piece from an earlier blog entry  on this very site:
Transition 16355T appears in subclade L5a, L2c, L2b, L2e , L1c3a1b, L3k1 , L4b2  and L2d . It’s worth noting the presence of this polymorphism in the so-called L-type aforementioned clades, but also, that while it appears in the R sub-haplogroup of the L3N clade, the location of both transition 16126C and 16355T in 2 mutually independent sub-haplogroups of the R clade, which are in turn mutually independent of hg N sub-haplo-groups N1a1a and N11 , where 16355T again appears, whereas either polymorphism is rendered absent in other sub-haplogroups of hg R and hg N super-clades, suggests that these polymorphisms have independently emerged multiple times in distinct mtDNA organelles.
Yet another potential example, this time from Barnabas et al...
The single mutation resulting in combined gain of 16389 HinfI and loss of 16390 AvaII sites, though variable, appears to be ancient and is one of the characteristics of African haplogroup L2. In the present sample set this haplotype is found in both ‘M’ and ‘N’ clusters (Figure 2).
The motifs associated with this sort of single nucleotide polymorphism may well be variable, as opposed to nucleotide similarities invoked other cases, but the resultant effect of said polymorphism is the same, speaking to that aforementioned prospect of convergence in evolution.
Rather than confronting the molecular-driven decision undertaken by Barnabas et al., particularly with regards to the association of the African haplotypes with the F clade, their ideologue detractors gratuitously dismiss all content of their work on the mere date of publication, which is branded as "aged", and therefore, rendering the work "outdated". At most, Chen et al., whom Barnabas et al. reference, as they utilize their data for the latter's study, are positioned as authority over Barnabas et al., without actually substantiating what it is that the former is supposed to be an authority over.
Reaction reminiscent of that directed at the Barnabas et al. report had also surfaced against the Quintana-Murci et al. (2004)  report on the M1 clade. Like that afforded to Barnabas et al., Quintana-Murci et al.'s detractors dismiss the contents of that report on the grounds of its "date of publication", and hence, "outdated". Meanwhile, molecular information revealed in that report, which has surfaced on this very site and referred to a few times (click, for details of said molecular specifics), have gone unaddressed and unchallenged by way of actual meaningful substance.
As in the case of Barnabas et al., Quintana-Murci et al.'s (2004) detractors even go as far as to suggest that the researchers fabricated molecular information, with virtually no rational explanation thereof, on occasion when confronted with DNA they have no other way of getting around. Characteristic restriction sites of the M clade, like np 10394 and np 10397 for instance, have been greeted with such helpless reactions, as comments on this very site would attest to, and again, reminiscent of reactions to the restriction site deletion at np 12406 in African L3 haplotypes, which has been treated as a characteristic of the F clade.
Even when an African gene pool deviates from the typological characterization of African gene pool by research teams, peculiarity and uniqueness (e.g. as in endemic) of such gene pool are hardly ever minded to any considerable degree, so as to open up to the possibility that preconceived notions about African phylogeny, and possibly human phylogeny overall, need to be reworked, and necessarily from ground up. More often than not, such occasion instead presents an opportunity for research teams to make a case for an exotic origin.
Such has been the case with such groups like "Nilo-Saharan" speakers in western and central Africa, and to some degree, those in eastern Africa, nomadic "Niger-Congo" speaking pastoralists (like the Fulani), or "Afro-Asiatic" speakers in "sub-Saharan" east Africa and coastal northern Africa, who turn up a genetic profile that doesn't neatly fit the African "prototype" profiles conceived by research teams. Example of this has been noticed in say, the lactase persistent genotype in pastoralist African ethnic groups, which have casually been dismissed as products of gene flow from outside Africa.
If the haplotype profile fits that seen outside Africa, so it appears, it is simply easier to attribute its presence in Africa to an outside source. Yet DNA analysis had brought to light, unique markers that are endemic to mainly pastoralist African ethnic groups, thereby demonstrating that lactase persistent trait in African pastoralist groups is primarily the product of local evolution to the lifestyle.
The finding of "lactase persistence" genotype in say, the west African Fulani, that happen to resemble those outside Africa, are casually treated as proof of a "non-African" origin in their ancestry, as if to say that without such gene flow, said pastoralists would have to have been lactose intolerant ("lactase non-persistent"). What else is there to interpret from such a theory? The comical nature of that very theory rarely crosses the minds of advocates thereof.
Whereas it has been argued on this site, that convergent evolution in lactase persistence-associated DNA cannot indeed be ruled out. Such a perspective takes into consideration, the prospect of the lactase persistent trait being an adaptive biological feature of pastoralist groups regardless of gene flow from an external source; meaning, such groups will more than likely develop lactase persistence, even if no foreign-sourced gene flow entered their gene pool. It's not like this phenotype is a disease-resisting deficiency; it's a response to a lifestyle.
Aside from the Fulani, the identification of lactase persistence genotype in eastern Africa, as the 13915G allele perhaps exemplifies, has been flaunted within ideological circles as a marker for "non-African" ancestry, even though eastern Africa is at the top of the list, where diversity in lactase persistent genotype is highest, featuring clades that are rare outside African pastoralists .
Maternally-mediated genotype has also been afflicted by the desire of research teams to make genotypes comply with their bipartisan framework of human ancestry, namely "African" and "non-African". Again, when an African gene pool deviates from the perceived African "prototype", it just seems convenient to simply place into "non-African" ancestry, and thereby simplify what could be part of a very complex picture, no matter how anomalous of a profile the corresponding genotypes may be to the preferred narrative[6, 7]. Consider the talked about mtDNA genetic profiles of Ethiopians, for example.
Even though there was a substantial enough segment of the Ethiopian mitochondrial gene pool, which apparently does not conform to what a good amount of research teams take as "African" ancestry, to warrant caution against implicating a South Arabian origin, certain research teams were not deterred in forcing said gene pool into their two-class framework ("African" vs "non-African"), so that it can comply with the prevailing arrangement of having any lineage that is not designated an "L" clade to be automatically rendered "non-African".
A "non-African" classification is a desirable arrangement, since it serves as a platform for linking said ancestry to South Arabia, notwithstanding the genotypic indicators suggesting differently. A strong desire in placing the genesis of the Semitic sub-phylum squarely across the Red Sea from Africa would mean getting around contradicting DNA revelations, and hence, ultimately arriving at a South Arabian origin point, which is where Ethio-Semitic would have to come, given allegations about Ethio-Semitic being phylogenetic closest to Old South Arabian languages.
To this end, the ancient Sabean complex of approximately 3000 years ago makes an attractive candidate as a likely source. Since some have been compelled to confront conflicting genetic information, which just doesn't seem to go away, settling for an origin in the Levant, where there is relatively more genetic correspondence than with South Arabia, is taken as the next best choice to that of an outright South Arabian origin. It is so argued, that ancient South Arabians would have had to come from the Levant, before settling South Arabia, thereby raising the prospect of ancient South Arabians being more genetically similar to Levantine populations than contemporary South Arabians.
Relevant quotes of the past :
The other two episodes of intrusion of Semitic influence, related to contacts with southern Arabia, are weakly supported by our data.
This is because, among the haplogroup N lineages present in high frequency in the Tigrais and other Ethiopian ethnic groups, only a few revealed close relationships with equivalent lineages present in southern Arabia. - Kivisild et al., 2004 
The uneasy correspondence with South Arabia, would tacitly mean that population replacement must be primarily responsible for contemporary South Arabian demography; alas, devout followers of a South Arabian origin are compelled to scratch their heads for requested evidence to that end. To make matters more difficult, as argued here before, the Ethiopian maternal gene pool has even better correspondence with surrounding Africans or Africans to their north than they do the Levant; take for instance, this from past postings :
...A specific haplotype match in haplogroup (preHV)1—which is also widely spread in the Near East—between Ethiopian Jews and non-Jews is more problematic, because it is also possible that the non-Jews obtained the lineage from the Jews. This particular (preHV)1 haplotype, with a rare transversion at np 16305, (1) has not been detected, so far, among other Semitic populations of the Near East; whereas, (2) in Ethiopia, it occurs both among Cushitic and Semitic speakers;
and, (3) in Ethiopian Jews, there are many sub-Saharan African lineages from haplogroups L0–L3. It is more likely, therefore, that the matching haplotype does not represent the incursion of Jewish maternal lineages into the Ethiopian gene pool but that this haplotype instead substantiates the extent of Ethiopian admixture in the Falasha population.
Taken together, the influx of the elements of the Hebraic culture in the first centuries a.d. probably did not have a major impact on the genetic pool of Ethiopians,
and the present-day Jews of Ethiopian descent probably assimilated genes from the local non-Jewish populations through conversion of the latter to Judaism.
Importantly, as already cited above...
The other two episodes of intrusion of Semitic influence, related to contacts with southern Arabia, are weakly supported by our data.
This is because, among the haplogroup N lineages present in high frequency in the Tigrais and other Ethiopian ethnic groups, only a few revealed close relationships with equivalent lineages present in southern Arabia.
Such unique associations should serve as a precautionary note against blanket dismissal of the gene pool as "non-African", but sadly, minds clouded by ideological thinking are not to be deterred in the pursuit of the bearers' own self-interest objectives. The African-specific X1 clade (largely restricted to "north Africa" and east Africa) may be serving as an example, among many, of this  cautionary notifier:
Two Ethiopian haplogroup X sequences from this study have been characterized elsewhere as belonging to North and East African–specific subclade X1 (Reidla et al. 2003). A control-region sequence similar to the Tigrai X1 haplotype was found recently in a Gurna sample from Egypt, though it was probably mislabeled as “L3” by the authors, since no coding-region markers specific to either haplogroup X or L3 were determined in that study (Stevanovitch et al. 2004). Both Yemeni X sequences, in contrast, belong to the major western Eurasian subclade X2. 
The near identical incidences, aside from peculiarities noted above, of what some take for granted as "non-African" ancestry in the Ethiopian maternal gene pool suggests a population history of considerably deeper time depth than the 3ky ago scenario idolized by followers of a South Arabian origin; their rationale is that subsequent inter-ethnic group intermingling must have evened out the distribution between the major Ethiopian groups. However, the failure of this trend to continue into the Y-DNA [6, 9] counterpart (in what some take for granted as "non-African"), where there seems to be a lingering genetic-structuring along linguistic lines, punches a hole into that frame of mindset.
If anything, the contrasting Y-DNA and mtDNA patterns, the former being largely one of linguistic variation, while the latter largely geographical variation, suggests different demographic processes being the likely driver. The mtDNA may well speak to a common ancestral gene pool from which the major Ethiopian groups emerged, while the Y-DNA is likely speaking to a subsequent differentiation attained in a group whose language was more Semitic-like in its fundamental features. This differentiation would then have had a temporal and spatial component to it [6, 9].
In the spatial component, the ancestors of the Cushitic speaking group would have been located in an area where say, hg J frequency was not as high as in that of the Semitic speaking branch. However, in line with mtDNA, Y-DNA too speaks to a deeper, prehistoric, time frame which discredits a 3ky ago time frame scenario [6, 9].
While on the subject of Y-DNA, why not revisit a case in the west African context, wherein this theme of slapping a "non-African" label onto ancestry surfaces, to cater to what appears to be some degree of self-censorship applied to clade assignment within DNA research team circles, and on the account of a clade not fitting neatly into what have become stereotypical renderings of what is supposedly "African"? For this, consider the unique group of hg R clades located mostly in western and central Africa, although some incidences have been noted in northeastern Africa as well.
A good segment of the clades in question are the R1b1c group (formerly designated as "R1b1a"). This group is particularly characterized by the V88 polymorphism. While this clade surfaces elsewhere, the overwhelming majority have been located in Africa.
The very few clades found outside Africa are generally examples that don't match the African examples, with one exception , the sharing of "undifferentiated" V88 chromosomes. Only two such chromosomes was identified outside Africa: one in a large (n = 465) "composite" sample from "western Asia", and one in another large (n = 510) "composite" sample, from the "Balkan" region.
Needless to say, the sharing of "undifferentiated" chromosomes between African and non-African samples, however rare it may be in the latter, is not necessarily a reflection of the same clades. The researchers had to sift through noticeably larger "non-African" samples just to get to any trace of the R1b1c clade, while the considerably lower sample sizes from Africa were enough to yield substantial amounts of R1b1c chromosomes.
Just to get an idea of the above, for example, consider the sample size of 1173 from Italy that was needed just to get any trace of R1b1c (just one chromosome), versus say, samples sizes from just 4 to 9 in several African samples that was required to get substantial amounts of R1b1c chromosomes, with the overall African R1b1c frequency ranging from 3 to 96 percent! To serve as another example for distribution outside Africa, only one R1b1c chromosome was found in a French sample from Corsica, presenting the highest frequency of R1b1c (at a mere 0.7%) chromosome anywhere outside Africa.
Clearly from the above notes, the likelihood of the African R1b1c chromosomes coming from Europe or elsewhere is very slim, as the African examples don't appear to be a subset of R1b1c chromosomes elsewhere. Hence, R1b1c having an African origin is a very real possibility.
The African distribution of R clades is not limited to R1b1c either. Hassan et al. (2008)  found that the R1 chromosomes found in their African (Fulani) samples, happened to be examples lacking the P25 mutation, that is a regular feature of R1b chromosomes found in Europe. Whatever may be said of the stability of P25, it is of note that an entire sample of R1 chromosomes lacked P25. By contrast, virtually all the R1b chromosomes found in Cruciani et al.'s  and Berniell-Lee et al.'s (2009)  samples tested positive for P25 [18, 19].
Were Hassan et al. alone in their observation? Think again! Wood et al.'s (2005)  report also pointed to the presence of R1 chromosomes lacking P25, but downstream of the M207 mutational event. Their African R chromosomes also extended to chromosomes that bore M207 but lacked any known downstream hg R mutations, which would make African samples among the unique samples in the world where the most basic R clades had thus far been identified!
As noted on this site before [18, 19], Wood et al.'s genotyping had accounted for all the important basic nodes of the R1 clade, just in case critics cry "low resolution" as the culprit behind the results they reported in their paper, with regards to hg R chromosomes. Perhaps an even more resolute genotyping procedure may reveal something entirely new that is yet to be understood about these rare African chromosomes, but the authors' genotyping was sufficiently resolute to remove the African chromosomes from those found elsewhere.
Arredi et al. (2004)  furthermore pointed to the presence of "undifferentiated" P clades in their African samples, which is ancestral to hg R, among others. Their genotyping resolution too appeared to be sufficiently resolute enough to disassociate said P chromosomes from those bearing well established P clade markers.
If these clades are indeed basic P chromosomes, then that opens up the possibility that not only were P clades present on the African continent at some point, which was explored on this site before , but also that an African origin for the R clade is also possible. Needless to say, this is an idea that is not very popular among Eurocentric ideologues or those who like to stereotype "African" to mean anything mainly outside of F-M89 (with the exception being the YAP+ hg E sibling clade of hg D) in Y-DNA phylogeny or anything outside of hg M and hg N clades in mtDNA phylogeny. This group of people will thus make emotional points one way or another about how "far out there" such an idea supposedly is, when indeed there is a solid objective groundwork for such an idea.
In what was perhaps a rare occasion in their study, Kivisild & co. did not gratuitously write an "L" designated clade off as "African", although the info that they leave with the reader, could very well render the homeland of the clade either way—either east African ("African"), or south Arabian, which the authors would more than likely treat as "non-African"; to recall, they note:
Third, the high frequency of haplogroup L6 in Yemenis points to an enigmatic link between the southwestern Arabian gene pool with that of East Africa. This haplogroup derives from the phylogenetic tree of sub-Saharan African mtDNA haplogroups but shows only marginal incidence in Ethiopians and is completely absent elsewhere in Africa.
Its high frequency in Yemen, together with low haplotype diversity, probably reflects the effect of genetic drift in a small founding population. A recent bottleneck of the general Yemeni population seems unlikely because of the high haplotype variation in other haplogroups (table 3).
A founder effect from outside is also not supported, because of the lack of a possible source population outside Yemen, in whom the L6 founder haplotype would be present at a significant frequency.
From the present evidence, the possibility cannot be eliminated that this haplogroup may even have originated from the same out-of-Africa migration that carried haplogroups M and N and founded the mtDNA diversity of Eurasia, the Americas, and Oceania. Yet, this scenario would imply a total isolation of a southern Arabian population from the others in that region to explain the absence of L6 types in other populations of the Near East, Arabia, and elsewhere in the world.
Alternatively, in consideration of the highly heterogeneous haplogroup composition of individual populations from East Africa (e.g., from Tanzania [Knight et al. 2003]) and the almost complete lack of data from some regions (like Somalia and Kenya), it is possible that the source population of Yemeni L6 varieties has not yet been sampled. 
The authors above actually open themselves up to the possibility that an "L" designated clade outside Africa could very well be a telltale sign of a local evolution from the often invoked "out-of-Africa migration" (OOA), as opposed to necessarily a reflection of post-OOA gene flow from Africa! This would naturally be at odds with ideologues who swear that "L" clades could not have accompanied OOA migration talked about above, on the account of "L" clades being generally rare in much of the non-African populace. Again, there is hilarity in how the underlying irrationality of that mindset eludes the consciousness of its proponents, for M and N could not have emerged even outside of Africa, without being preceded by their ancestral L3 clades.
Kivisild and co.'s notes above is reminiscent of Barnabas et al.'s approach to exploring F, M & N clades origins, or say, Gonder et al.'s (2007)  opening themselves up to the possibility of the Tanzanian M clades reflecting remnants of the African source for such lineage, which serve as relatively rare occasions which refreshingly stand out from the seemingly traditional habits applied in many research papers, which are suggestive of tunnel-vision-like perception.
Coastal north African populations, particularly those in the Maghreb, are frequently portrayed by ideologues as long-term settler transplants of Europeans/Asians on the African continent, on the account of those ancestry that are supposed to be "non-African", usually with little to no regard to primary "African" ancestry as inferred from the paternal genotype [11, 20], not leaving out a language which essentially defines the Amazigh groups, which is entirely unique to the African continent. Mappings have shown time and again, however, that far from being mere Eurasian transplants, coastal north African demography has its own peculiar genetic profile that lends it the inclination to position between African samples and European or Asian samples .
The peculiarity of the positioning of coastal north African specimens, as pointed out , is not only a reflection of some shared attributes to continents on either side of the Mediterranean sea, but also of ancestry unique to coastal north Africans (as elements of atDNA  indicate), which itself may be manifestations of an ultimate proto-Tamazight (proto-"Berber") ancestry in the vicinity of eastern Africa [11, 20, 13, 22]. The peculiarities of coastal north African maternal gene pool, some of which were exemplified above in tandem with notes on Ethiopian/east African gene pool, echo this. To add to this, U6 and basal M1 clades perhaps stand as among the often-cited examples of such clades, where the Maghreb is concerned.
The generally gratuitously-monotypic characterization of coastal northwest African gene pool—usually in the form of how these Africans are supposedly more related to some "Eurasian" or the other than fellow Africans—came under a spotlight, for example, whereby Cherni et al. (2005)  were taken aback when their results blew away the expectations they had coming into the study. This experience certainly had said authors reflect on procedural tendencies that afflict samplings of coastal north Africa and the resultant casualties in emerging imagery painted about the northwest African landscape.
It would be one thing if by "non-African", the term seeks to emphasize ancestry that doesn't necessarily preclude "African origin", but to stress the "transcendence" of ancestry in Africa; however, many research teams, particularly those in the "west", tend to use the term to mean ancestry "exclusive" of African ancestry. This is where the problem lies, as human phylogeny is not so simple a picture as research teams would prefer it be.
The notes in the discussion segment seek to emphasize certain peculiarities about African DNA, be they Y-DNA, atDNA or mtDNA, particularly those that do not comply with the description of the stereotypical African clades, which is also specified above, and hence, gratuitously ascribed to "non-African" ancestry. The idea is to instill caution against such premature, if not primarily ideological, classification of those African clades—which do not conform to the descriptions ascribed to the stereotyped African clades—into "non-African" ancestry.
In the discussion segment, several specific examples had been recalled to drive home this point. In a number of cases, doubt was raised about the African clades being derivatives of the so-called "non-African" counterparts, and hence, put into question the prospect of a "non-African" origin for such clades, even though a good amount of people have this seemingly unbreakable habit of calling these "non-African". This is not to say that elements of what said people call "non-African" cannot be a reflection of the so-called "back-migration," given that humans do move about, and there is always a good chance that geographically proximate populations or groups will exchange genes to some degree or another, notwithstanding the occasional barriers presented by cultural differences. Rather, it's simplistic and erroneous to assume that all so-called "non-African" clades in African populations are well, reflective of a "non-African" origin.
As several blog entries have indicated, as does the discussion segment right above, a picture starts to emerge whereby it appears that a good chunk of what is dismissed as "non-African" in African samples, may actually reflect ancestry that was on the African continent all along, and may possibly be reflecting a segment of the African population that went onto populate other locations of the globe outside the African mainland. The Sahara appears to be one spot where a good segment of what is called "non-African" ancestry had considerable presence, and not necessarily because "non-Africans" settled therein!
This may explain a reasonable amount (obviously not all) of so-called "non-African" clades found in Maghrebi populations, thereby giving the genetic landscape of that region the impression of an asymmetric "African" (primarily paternal) and "non-African" (mainly maternal) ancestry, as well as nomadic pastoral groups from western Africa and such populations now living in the African Horn, like say, Semitic and Cushitic segments of the populace, or even groups in southeastern Africa. Sub-Saharan populations like the latter (certain Ethiopian groups) may reflect a biohistory whereby their prehistoric ancestors, or at least a portion of them, were once living in a more northerly latitude in the Sahara, where many of the so-called 'non-African" clades would have been situated.
As such, changing climatic conditions from favorable to adverse may have urged groups to seek refuge in the more climatically stable, and possibly more favorable, areas of the African continent. Some migrants built communities around persisting river systems, as the development of ancient Nile Valley complexes perhaps exemplify, while others proceeded to the more vegetated areas of the equatorial region and thereof.
The sharing of supposed "non-African" clades of Ethiopian groups more immediately with coastal north Africans than those from the Levant, and sharing even less so, with those from south Arabia, which is where Ethio-Semitic languages would have to come from, i.e. if they came from outside Africa, points to this scenario just now described. So-called "non-African" clades are presently scattered across the continent, from west to east, and north to south, although noticeably unevenly; the sporadic appearances in such places like equatorial western Africa may be a reflection of either near population replacement through subsequent demographic processes or the region could have been relatively peripheral to the occupation of autochthonous African groups who carried supposed "non-African clades".
Low incidences in regions well beneath the equator, like southern Africa, on the other hand, more than likely reflect that aforementioned originally more northerly location of groups carrying the supposed "non-African" clades. While it's possible M1 emerged south of the equator before diversifying north of it, as the southeast African M clade distribution might indicate [3, 4,12], there is also a very good indication that the reverse, i.e. M1 emerging in the equatorial region or slightly north thereof [3, 4,12, 23, 24, 25], and then diffusing to the south, as result of said climatic changes in the Sahara serving as a pump for human dispersal within the African continent, explains the current distribution of M1 across the continent, given the genotypic specifics that have frequently been the subject of this site.
Several nucleotide information, as cited in the discussion segment, open up the possibility of an African origin for the very clades that are termed as "non-African"; this has been seen in the case of the F clade (i.e. sub-clade of N), and the M and N clades which serve as the main parental mtDNA clades of "non-African" gene pool. Certain African haplotypes have been featured by their similarities, yet more basic in phylogenetic status, to the M and N clades. The African haplotypes like say, AF24, 034 and 032 (M-like) or the AF7, AF15, AF21 and AF10 (N-, and F-like in some cases) are examples.
In short, Africa has all the necessary clades in place to make a case for the African origin of these so-called "non-African" clades; contrast this with the prevailing situation outside Africa. This makes sense, considering humans had spent the majority of their evolutionary lifespan in Africa. Yet somehow some are astonished or dumbfounded at the very prospect of Africa being the actual location of origin for supposed "non-African" clades.
Much of the time, perhaps more so in the mtDNA network—which is relatively larger and more complex, the ancestral clades for "non-African" clades are elusive, and are only inferred from hypothetical reconstructions using the phylogeny of preexisting haplogroups, as opposed to actual tangible clades that are suggestive of a basic/ancestral phylogenetic status. On the flip-side, there is some indication that OOA migration(s), leading to major peopling of other parts of the world, would have taken place in accompaniment of some "L" designated African clades, with these clades having largely undergone negative genetic drift in most places outside Africa, thereby explaining the low or negligible frequencies. Such clades as the L6 in Yemeni may perhaps be reflective of this, as one of two logical possibilities.
Perhaps the act of making say, clades M and N, the exclusive domain of "non-African" ancestry serves an emotional purpose, so that "non-Africans" can be given a sizable measure of importance in human phylogeny. However, to reiterate, human phylogeny is not so simplistic a picture in its actuality. The prospect of having a lion's share in the origin of human phylogeny is inclined to biased in the direction of Africa than "non-Africa"; acknowledging this doesn't speak to someone's nationalistic-driven "Afrocentric bias", but rather, just coming to terms with the aforementioned fact that humanity spent the bulk of its biohistory in the African continent, and that the human phylogeny reflects this, amongst other things.
Subject to revision or modification where necessary, upon reception of new or additional information.
: Pagani et al., Ethiopian Genetic Diversity Reveals Linguistic Stratification and Complex Influences on the Ethiopian Gene Pool, 2012.
: Blog entry, May 2013, What Ethiopian Genetic Diversity—Really—Reveals!
: Barnabas et al., High-Resolution mtDNA Studies of the Indian Population: Implications for Palaeolithic Settlement of the Indian Subcontinent, 2005.
: Quintana-Murci et al., Where West Meets East: The Complex mtDNA Landscape of the Southwest and Central Asian Corridor, 2004.
: Tishkoff et al, Convergent adaptation of human lactase persistence in Africa and Europe, 2007.
: Blog entry, January 2010, Following Trails of the Cro-Magnon - I & II
: Blog entry, May 2013, What Ethiopian Genetic Diversity—Really—Reveals!
: Kivisild et al., Ethiopian Mitochondrial DNA Heritage: Tracking Gene Flow Across and Around the Gate of Tears, 2004.
: Blog entry, September 2009, E-M34: Designation as "African" presents a Dilemma?
: Blog entry, December 2011, Genetic Profile: Maghrebi populations are Not "Satellite-Europeans.
: Blog entry, January 2008, RFLPs: Lucotte et al., A case study Pt1 & Pt2.
: Gonder et al., Whole-mtDNA Genome Sequence Analysis of Ancient African Lineages, 2007.
: Cherni et al., Female gene pools of Berber and Arab neighboring communities in central Tunisia: microstructure of mtDNA variation in North Africa, 2005.
: Cruciani et al., Human Y chromosome haplogroup R-V88: a paternal genetic record of early mid Holocene trans-Saharan connections and the spread of Chadic languages, 2010.
: Hassan et al. 2008, Y-chromosome variation among Sudanese: Restricted gene flow, concordance with language, geography, and history.
: Berniell-Lee et al., Genetic and Demographic Implications of the Bantu Expansion: Insights from Human Paternal Lineages, 2009.
: Wood et al., Contrasting patterns of Y chromosome and mtDNA variation in Africa: evidence for sex-biased demographic processes, 2005.
: Blog entry, January 2008, R1*-M173 Chromosomes in Africa.
: Blog entry, December 2009, R1*-M173 Chromosomes in Africa - II.
: Arredi et al., A Predominantly Neolithic Origin for Y-Chromosomal DNA Variation in North Africa, 2004.
: Blog entry, March 2009, Working hypothesis around haplogroups IJK, I, J, K, P, R, and Q.
: Blog entry, January 2009, Unwinding the Convoluted Character of the Emergence of Imazighen Groups.
: Gonzalez et al., Mitochondrial lineage M1 traces an early human backflow to Africa, 2007.
: Blog entry, January 2008, Mitochondrial DNA M1 haplogroup: A Response To Ana M. Gonzalez et al. 2007.
: Blog entry, April 2011, Summarizing clade M1.
: Pala et al., Mitochondrial DNA Signals of Late Glacial Recolonization of Europe from Near Eastern Refugia, 2012.
: Blog entry, April 2013, Update on Investigation into the "Mysterious" EpiPaleolithic Maghrebi Remains!
—Personal notes, 2005-2013.